Biomedical Engineering Reference
In-Depth Information
Chapter 2) for all the factors involved in the calculation. Thus, for the
upper-bound calculation: the aperture is made larger by an amount
intended to characterize patient and organ motion and registration
errors; the CT densities are reduced by an amount intended to
characterize the uncertainty in CT number; the dose is raised
everywhere by an amount intended to characterize the possible
variations in dose monitoring and calibration; and so forth. The
“lower-bound” dose calculation uses the opposite extremes.
These three dose calculations allow one to quote a dose with un-
certainty bounds at every point within the patient. The distribu-
tions should be interpreted with caution since the uncertainties
at different points are highly correlated. As a consequence, neither
the upper nor the lower-bound dose distribution is physically possible
and the displays tend to overestimate the amount of tissue with-
in which there might be a problem. Nevertheless, this approach
provides a crude estimate of uncertainty that can be useful in warning
of possible problems and can lead to a search for more “robust”
solutions whose uncertainty bounds are smaller.
Display of uncertainty
Figure 8.9. Coronal sections of a plan which, due to the
large field sizes involved, required abutting superior and
inferior fields. (a) Nominal dose distribution, (b) upper
bound dose distribution, (c) lower bound dose distribution,
and (d) upper bound dose distribution when the junction
between the fields is feathered (85% confidence limits.)
Reproduced with permission from Urie
et al
. (1991).
